There have been proposed a variety of photovoltaic elements for solar cells and for power sources in electric appliances. They utilize the pn junction formed by ion implantation or thermal diffusion of impurities into a substrate of single crystal of silicon (Si) or gallium arsenide (GaAs), or by epitaxial growth of an impurity-doped layer on a substrate of such single crystal. However, there is a disadvantage for these photovoltaic elements that their production cost unavoidably becomes high because of using a single crystal substrate. Because of this, they have not yet gained general acceptance for use as a solar cell or as a power source in electric appliances.
Recently, there has been proposed a photovoltaic element in which there is utilized a pin junction of amorphous silicon (hereinafter referred to as "A-Si") deposited film formed on an inexpensive substrate of non-single crystal material such as glass, metal, ceramics, and synthetic resin by way of the glow discharge decomposition method. This photovoltaic element has a nearly satisfactory performance and is of low production cost and because of this, it has been recognized as usable as a power source for some kinds of appliances such as electronic calculators and wrist watches.
However, for this photovoltaic element, there is a disadvantage that the output voltage is low because the band gap of the A-Si film constituting the element is 1.7 eV, which is not large enough. There is another disadvantage that its photoelectric conversion efficiency is low for a light source such as fluorescent light which contains more short-wavelength light, so that its application is limited to appliances with very small power consumption.
There is a further disadvantage for said photovoltaic element that the constituent A-Si film is often affected by the so-called Staebler-Wronski effect, in which the film characteristics are deteriorated upon continuous irradiation with intense light for a long period of time.
For a photovoltaic element to be utilized as a power supplying solar cell, it is necessary to convert efficiently and continuously the light energy of sunlight into electric energy, and hence, it is desired to have such a layer structure that permits photoelectric conversion for sunlight over as broad a spectrum range as possible.
Now, in the case where a photovoltaic element is made of a semiconductor material having a small band gap energy, the wavelength region of light to be absorbed by the layer will be extended from the short wavelength side to the long wavelength side. However, in this case, it is the long-wavelength component of sunlight alone that contributes to photoelectric conversion, and the energy of the short-wavelength component is not utilized for photoelectric conversion. This is because the amount of energy to be taken out by the photoelectric conversion is determined by the band gap energy of the semiconductor material as used.
On the other hand, in the case where a photovoltaic element is made of a semiconductor material with a large band gap energy, the wavelength component which is absorbed by the layer and contributes to photoelectric conversion is the short wavelength light having an energy greater than the band gap energy of the semiconductor material, and the long-wavelength component is not utilized for photoelectric conversion.
In a photovoltaic element, the maximum voltage or open-circuit voltage (Voc) to be supplied is determined by the band gap energy of the semiconductor materials utilized. In order to obtain a high Voc, semiconductor materials having a large band gap energy are desirable.
Therefore, there is eventually a limit for the photoelectric conversion efficiency for a photovoltaic element prepared by using a single semiconductor material.
The foregoing led to the idea of forming a plurality of photovoltaic elements using a plurality of semiconductor materials each having a different band gap energy, so that the individual photovoltaic elements become responsible for utilizing the different wavelength regions of sunlight. This idea was expected to contribute to an improvement in the photoelectric conversion efficiency.
However, there is a disadvantage for the solar cell having such a structure as mentioned above in that overall high photoelectric conversion is possible only in the case where the individual photovoltaic elements have good characteristics, because it is of such structure that a plurality of photovoltaic elements are stacked to form an electrically serial structure
Unfortunately, for the photovoltaic element having the foregoing structure, there has not yet been realized any desirable one wherein the respective constituent elements as stacked have satisfactory values of band gap energy and satisfactory characteristics as desired and provides a high Voc as the photovoltaic element.
There have been proposed direct transition-type semiconductor films having a wide band gap, such as ZnSe (having a band gap of 2.67 eV) and ZnTe (having a band gap of 2.26 eV) and mixed crystals thereof ZnSe.sub.1-x Te.sub.x (where 0&lt;x&lt;1). Public attention has been focused on these semiconductor films. These semiconductor films are, in general, formed on a substrate of single crystal by way of epitaxial growth. The as grown film of ZnSe exhibits n-type conductivity and the as-grown film of ZnTe exhibits p-type conductivity. However, for any of these films, it is generally recognized that it is difficult for the film to exhibit opposite type conductivity. Further, in order to carry out the epitaxial growth for the film formation, it is required to use a specific substrate of single crystal and to maintain the substrate at elevated temperature. And in this film formation, the deposition rate is low. Because of this, it is impossible to perform epitaxial growth on a commercially available substrate which is inexpensive and low heat-resistant such as glass and synthetic resin These factors make it difficult to develop practically applicable semiconductors films using the foregoing commercially available substrates.
Even in the case where a semiconductor film should be fortunately formed on such commercially available substrate, the film will be such that is usable only in very limited applications.
In fact, there have been various proposals to form a direct transition-type semiconductor film on a non-single crystal substrate such as glass, metal, ceramics and synthetic resin. However, under any of such proposals, it is difficult to obtain a desired direct transition-type semiconductor film having satisfactory electrical characteristics because the resulting film is accompanied with defects of various kinds which make the film poor in electrical characteristics and on account of this, it is difficult for the film to be controlled by doping it with an impurity.
In the meantime, an amorphous film composed of Zn and Se elements is described in U.S. Pat. No. 4,217,374 (called "literature 1"hereinafter) and also in U.S. Pat. No. 4,226,898 (called "literature 2"hereinafter). And ZnSe compound is described in Japanese Patent Laid-open No. 189649/1986 (called "literature 3"hereinafter) and Japanese Patent Laid-open No. 189650/1986 (called "literature 4"hereinafter).
Now, literature 1 discloses amorphous semiconductor films containing selenium (Se), zinc (Zn), hydrogen (H) and lithium (Li); but the principal subject is amorphous selenium semiconductor films, and the Zn described therein is merely an additive, as are Li and H. And as for the Zn and the Li, likewise in the case of the H, they are used for reducing the localized state density in the band gap energy without changing the inherent characteristics of the film. In other words, the addition of Zn to the amorphous Se film mentioned in literature 1 is not intended to positively form a ZnSe compound. Incidentally, literature 1 mentions nothing about the ZnSe compound and the formation of ZnSe crystal grains. Regarding the addition of Li, it should be noted that it is not added as a dopant.
Literature 2 does mention amorphous semiconductor films containing Se, Zn, and H. However, it deals mainly with amorphous silicon, and it defines Se as an element which forms a compound with said silicon. As for the Zn, it is defined as an element which sensitizes the photoconductivity and reduces the localized state density in the energy gap. In other words, the additions of Zn and Se are not intended to form a ZnSe compound. Incidentally, literature 2 mentions nothing about the ZnSe compound and the formation of ZnSe crystal grains.
Literature 3 and literature 4 are concerned with the deposition of a ZnSe film by the HR-CVD method (hydrogen radical assisted CVD method). That is, they disclose methods of improving the deposition rate and the productivity of a deposited film; but they merely mention deposited films of non-doped ZnSe.
Against this background, there is an increased social demand to provide an inexpensive photovoltaic element having a high photoelectric conversion efficiency, particularly, for short-wavelength light which may be practically usable as a solar cell and also as a power source in various electric appliances.